• Energy Research

Electrolyte imbalance caused by the undesired vanadium-ions cross-over and water transport through the membrane is one of the main critical issues of vanadium redox flow batteries, leading to battery capacity loss and electrolytes volume variation. In this work, the evolution of discharged capacity and electrolyte volume variation were firstly investigated adopting commercial electrolyte for hundreds of charge-discharge cycles in vanadium redox flow batteries employing different membranes, varying thickness and equivalent weight. Subsequently, with the support of a 1D physics-based model, the origin of the main phenomena regulating capacity decay and volume variation has been identified and different modifications in the preparation of electrolytes have been proposed. Electrolytes characterized by an equal proton concentration between the two tanks at the beginning of cycling operation turned out to limit capacity decay, while increasing electrolyte proton concentration was effective also in the mitigation of volume variation. The most promising electrolyte preparation combined the effect of high proton concentration and null osmotic pressure gradient between the two tanks: compared to commercial electrolyte this preparation reduced the capacity decay from 47.7% to 20.9%, increased the coulombic efficiency from 96.2% to 98.9% and the energy one from 79.9% to 83.4%, and also implied a negligible volume variation during cycles. The effectiveness of this electrolyte preparation has been verified with different membranes, increasing the range of validity of the results, that could be thus applied in a real system regardless of the adopted membrane.

Vanadium redox flow battery (VRFB) is a very promising solution for large‐scale energy storage, but some technical issues need to be addressed. Crossover, i.e., the undesired permeation of vanadium ions through the cell separator, causes capacity loss and self‐discharge. Low‐cost and highly selective separators are thus required to improve the competitiveness of this technology. This work investigates the use of silica nanoparticles in an innovative selective layer to improve membrane selectivity and reduce its thickness. 1.5 μm thick barrier layers composed of 1100EW Nafion ionomer with silica (≈3–13 nm diameter) and Vulcan XC‐72R (≈40 nm) nanoparticles in different proportions are directly deposited on 50 μm thick Nafion membranes. The barrier layer composed only of silica nanoparticles reduces the self‐discharge due to crossover by 5 times and increases the average efficiency of the battery. Finally, during more than 1000 h of operation, the barrier layer on a 25 μm Nafion membrane demonstrates excellent stability, working with a constant coulombic efficiency higher than 99% and a capacity decay rate comparable with a thicker Nafion membrane, thus enabling the use of thinner membranes in VRFB, allowing an estimated 8% stack costs reduction with respect to NR212.

Abstract Water transport control is the major issue of direct methanol fuel cell. High water flow rates through the fuel cell imply extra water feeding at the anode and flooding at the cathode. In the present work, water transport and flooding in the direct methanol fuel cell are investigated through both experimental and modeling analyses and an interpretation of such phenomena is proposed. The model is validated on the experimental data of two different fuel cells in an extensive range of operating conditions. The analysis elucidates that water transport through the cathode diffusion layer is determined by vapor diffusion, slightly affected by current density, and by liquid water permeation proportional to current density, that occurs when liquid pressure in the electrode exceeds a threshold value. To simulate the effects of cathode diffusion layer flooding two mechanisms must be considered simultaneously: superficial pore obstruction, proportional to liquid water concentration in cathode channel, and bulk pore obstruction, proportional to liquid water permeation. The modeling analysis proposes the correlations to reproduce the effects of cathode flooding and permits to discuss the onset and magnitude of such phenomenon and the influence of micro-porous layer.

Abstract Polymer electrolyte membrane fuel cells are devices that produce power by direct conversion of hydrogen via electrochemical route and are promising for energy applications, mainly because no direct pollutants are produced during operation. Automotive is the major industrial application for polymer fuel cells, which could replace internal combustion engines as power sources, conditionally to the achievement of a significant cost reduction. Increasing power density and reducing the loading of precious metal based catalysts is thus a technological priority. In this direction, the geometry of the flow field plays a dramatic role: at state of the art, hydrogen and oxygen are distributed over the fuel cell area through channels. Non-uniform distribution of reactants, which results from non-optimal flow field design, determines heterogeneity during operation, loss of efficiency and accelerates ageing. In this work, computational fluid dynamics is used to analyse oxygen transport in a low platinum polymer electrolyte fuel cell for automotive applications. Analysis focuses on the effect of 3D geometrical features that are present in state of the art flow fields. Comparison of three flow fields (straight channel, serpentine and interdigitated) is performed and it is observed that the contact points between the GDL and the current collector determine significant performance loss because of sluggish oxygen transport in these regions. Nevertheless, a trade off with electron transport through the GDL must be considered. To support the conclusions of the work, an original methodology is adopted, by simulating electrochemical impedance spectroscopy, an experimental transient technique that allows to selectively evidence the effect of mass transport from other physical phenomena.

This study combines local electrochemical diagnostics with ex situ analysis to investigate degradation mechanism associated to start-up/shut-down (SU/SD) of PEMFC and mitigation strategies adopted in automotive stacks. Local degradation resulting from repeated SU/SD was analyzed with and without mitigation strategies by means of a macro-segmented cell setup provided with Reference Hydrogen Electrodes (RHEs) at both anode and cathode to measure local electrodes potential and current. Accelerated Stress Test (AST) for start-up with and without mitigation strategies are proposed and validated. A ten-fold acceleration of performance loss due to un-mitigated SU/SD has been calculated with respect to AST for catalyst support. Under mitigated SU/SD, instead, a strong degradation was observed as localized at cathode inlet region (i.e. −38% ECSA loss and −22 mV voltage loss after 200 cycles) due to local potentials transient reaching up to 1.5 V vs RHE. These localized stress conditions were additionally reproduced in a zero-gradient and a new AST protocol (named start-up AST) was proposed to mimic the potential profile observed upon SU/SD cycling. Representativeness of the start-up AST for real world degradation was confirmed up to 200 cycles. Platinum dissolution and diffusion/precipitation within the polymer electrolyte was shown to be the dominant mechanism affecting performance loss.

A long-term dynamic load cycle is performed on state-of-the-art membrane electrode assemblies, aiming to evaluate the degradation mechanisms of Polymer Electrolyte Membrane Fuel Cell under real-world automotive operations. The load cycle, adapted from the stack protocol defined in H2020 ID-FAST European project, includes load, pressure and temperatures cycling. Events that recover the temporary decay are included, specifically procedures classified in short-stops, cold-soaks, long-stops. Operando voltage and current distribution are measured through a segmented hardware, combined to local in-situ electrochemical characterization. Investigation is supported by scanning and transmission electron microscopy analysis, performed at different locations along-the-flow-field. Reversible degradation weights from few to 20 mV and changes local current distribution, mostly at air-inlet, since the dry-out of ionomer. Cycle efficiency decreases of 3%-9%: the largest irreversible performance losses are observed at air-inlet, while middle-region is the least impacted. Cathode catalyst layer and membrane are the most aged components: platinum active surface area drops in 200-400 hours, because of electrochemical Ostwald ripening mechanism, and stabilizes around 62%-67% of initial value. Polymer membranes report ageing compatible with mechanical stress that causes localized thinning, increasing hydrogen crossover. Decay of ionomer in the catalyst layer is discussed, which would consistently explain alterations of mass transport resistance. International audience

Abstract In fuel cell based combined heat and power (CHP) plants, degradation within the fuel cell stack and the steam methane reformer significantly affects the generated electrical and thermal power. As a consequence, incorporating system’s degradation within the model of the plant could be of great importance in order to estimate the resulting variations in the electrical and thermal power generation and taking appropriate measures to mitigate such deviations. To this end, in the present article, a multi-objective optimization approach has been proposed and employed to find the optimal operating parameters of an HT-PEM fuel cell based micro-CHP system within the first 15,000 h of operation while considering the impact of degradation. Two different optimization procedures with the following objective functions have been applied: (I) net electrical efficiency and thermal generation; and (II) net electrical efficiency and electrical power generation. Steam to carbon ratio, auxiliary to process fuel ratio, fuel partialization level and anodic stoichiometric ratio are the design parameters. Based on the results of optimization procedure I, the highest achievable net electrical efficiency at the beginning of operation is 32.75% which, due to degradation, considerably declines to 29.51% in the last time interval. Moreover, in all time steps, optimal solutions cover a wide domain of thermal generation which assures the capability of the system to easily cope with the thermal demand of the user. On the other hand, optimization procedure II displays a steady decrease in both electrical efficiency and electrical generation through time which indicates the adverse effect of degradation on these two performance indices. Finally, it has been found that, using optimization procedure I, the cumulative average electrical efficiency of the plant improved from 26.03% at normal operation to 27.56% at optimized condition. Furthermore, it was determined that by employing the optimal points obtained in optimization procedure II, the average cumulative electrical power generation is increased from 25.4 kW (at normal operation) to 26.8 kW. It is noteworthy that, the study not only provides some insights into the long-term performance of such system, but can be more importantly perceived as a guideline to adaptively optimize the operating conditions of the system in order to alleviate the degradation’s effect and to guarantee optimal performance of the system throughout its lifetime.

The electrochemical impedance spectroscopy (EIS) characterization technique, although widely adopted in electrochemistry for understanding operational issues and degradation, has a less consolidated physical interpretation in lithium‐ion batteries (LIBs), often relying on circuital methods. Herein, the Doyle–Fuller–Newman model is adapted and experimentally validated for the physical simulation of electrochemical impedance; then, it is applied in a comprehensive one‐factor‐at‐time sensitivity analysis on an impedance spectrum from 4 kHz to 0.005 Hz; 28 physical parameters, which represent the kinetic, resistive, diffusive, and geometric characteristics of the battery, are varied within broad literature‐based ranges of values, for each of the 20 analyzed battery states, characterized by different state‐of‐charge and temperature values. The results show a miscellaneous sensitivity of parameters on impedance spectra, which ranges from highly sensitive to negligible, often resulting in a strong dependence on operating conditions and impedance frequency. Such results consolidate the understanding of LIB electrochemical impedance and demonstrate that 40% of the parameters, 12 out of 28, can be considered poorly sensitive or insensitive parameters; therefore, fitting the experimental EIS data, their value can be assumed from the literature without significantly losing accuracy.